This is a multicenter, double-blind, randomized comparison of two different medical treatments to prevent recurrence of duodenal ulcer (DU) and gastric ulcer (GU) bleeding. The overall HYPOTHESES TO BE TESTED are that 1) eradication of Helicobacter pylori (H. pylori) infection alone will cure most patients of their chronic peptic ulcers and will be substantially equivalent to H.pylori eradication plus full-dose H2RA maintenance in preventing recurrent ulcer hemorrhage and 2) ingestion of aspirin (ASA) or other non-steroidal anti-inflammatory drugs (NSAIDs) or recurrence of H.pylori infection will account for all recurrences of ulcer hemorrhage during long-term follow-up. To test these hypotheses, approximately 310 patients who had a clinically significant DU hemorrhage & 310 with GU hemorrhage will be randomized into this study after ulcer healing & eradication of H. pylori are documented. They will be followed for a median duration of at least 36 months. All patients without H.pylori eradication at the time of referral will be treated with a two- week course of antimicrobial therapy (Omeprazole 40 mg QD & Clarithromycin 500 mg TID). Patients in whom H. pylori is now eradicated, as determined by C13 urea breath testing 6 weeks later, will then be randomized to receive in a double-blind fashion either a full dose of a histamine-2-receptor antagonist (H2RA-Famotidine 40 mg HS) or placebo for the entire duration of the study. Patients whose H. pylori is not eradicated will be treated with another two-week course of antimicrobial therapy (Omeprazole 20 mg BID and Amoxicillin 1 gm BID or triple therapy if penicillin allergic). Patients in whom H.pylori is now eradicated will be randomized into the double-blind study. Those patients whose H.pylori is not eradicated will be treated with full-dose H2RA and followed long-term. SPECIFIC AIMS are 1) to determine the efficacy and safety of H.pylori eradication alone versus H.Pylori eradication combined with daily full dose H2RA in preventing recurrence of DU & GU hemorrhage; 2) to determine the recurrence rates of H.pylori infection during a median follow-up of 36 months; 3) to document whether recurrences of hemorrhage are associated with NSAIDs-ASA and/or H.pylori recurrence of H.pylori infection; 4) to compare the direct costs (unscheduled use of health care) & indirect costs (time lost from work or usual activity) associated with each treatment group. The PRIMARY OUTCOME is the incidence of recurrent bleeding from DU or GU with each treatment. SECONDARY OUTCOMES are frequencies of symptomatic non- bleeding ulcer, treatment failure, NSAID-ASA ingestion, re-infection with H.pylori and direct and indirect costs of care with each therapy. Survival analysis will be used in addition to other statistical methods. The SIGNIFICANCE OF THIS STUDY is that 1) new scientific information about prevention of a major complication of gastric and duodenal ulcers will result; 2) the efficacy, safety, & cost of H. pylori eradication (with and without continued H2RA) for patients with an GU and DU hemorrhage will be defined; and 3) these outcomes will have profound implications for the management of a large group of patients worldwide with complicated peptic ulcers.